The Flames have been out of action for nearly a month now. It’s not an easy pill to swallow seeing how wide open the playoffs have been and how the Flames aren’t competing anymore, but there’s still plenty of reason to feel good about the Flames.
Let’s not forget how good of a campaign they put together, considering they heading into the season with lowly expectations of just making the playoffs. They dominated more often than not, and racked up 50 wins.
In many of their wins and even their losses, Flames players were often picked as star selections at the conclusion of every game. I wanted to make a fun data visualisation that is representative of every single star selected and what their performances consisted of.
Star selections as data-driven badges
In a recent post by Elijah Meeks (@Elijah_Meeks) of the Data Visualization Society (DVS), he explored the reasoning behind what makes a data-driven badge so appealing. Setting out in a similar vein, I wanted to explore Flames and opponent star performances using visual badges.
For players, I selected goals, assists, and shots as the parameters of interest. For goaltenders, I selected shots against and save percentage. These metrics were chosen for the sake of simplicity. A more advanced approach towards using three star selections to make a game score model was created by Emmanuel Perry (@mannyelk) two years ago.
For the visualisation, games are grouped by 10-game segments, with the first, second, and third star being placed in order vertically. From a high-level perspective, the use of colour indicates whether a Flame was selected or if it was an opponent. The orientation of each triangle also shows if it was a skater or a goaltender.
The purpose of the visualisation was to create a fun way to revisit the Flames’ season. Perhaps in the future, the inclusion of more stats can be used to create a more comprehensive picture. In the meantime, please enjoy this as more of a data-art piece recapping the Flames’ season!
Below is the legend to read each badge. Note that in the full chart, the centre of the badges are highlighted with a white circle to give better context as to what the values are for each stat.
For players, high offensive outputs often warranted star selections, as seen by larger badges. Other times, additional on-ice performances and metrics such as blocks, hits, faceoffs, etc., went into the subjective selection of best players in a game. While those stats aren’t accounted for, exploring some of the smaller badges can reveal insight on what other aspects of a player’s game stood out.
Admittedly, the goaltender badges aren’t as easy to follow. It was difficult to scale things properly to match the general size of player badges, while having discrete visual boundaries between various goaltender performances, since more often than not, a goaltender selected as a star will have a high save percentage anyway.
Without further ado, here’s the entirety of the Flames’ season as visualised by three star selections.
Which games or performances stand out for you? Can you spot Johnny Gaudreau‘s six-point night? Matthew Tkachuk‘s first career hat trick? What about Sean Monahan‘s 10 shot night? The abundance of red really does remind us of how good the Flames were this past year.
What are your overall thoughts? Are there other stats worthy of inclusion or areas for improvement? Let me know in the comments or at @wincolumnblog.
Data from NHL.com. Chart made with Adobe Illustrator.